Swine influenza A virus infection dynamics and evolution in intensive pig production systems

Abstract

Swine influenza A virus (swIAV) is one of the main viral pathogens responsible for respiratory disease in farmed pigs. While outbreaks are often epidemic in nature, increasing reports suggest that continuous, endemic infection of herds is now common. The move towards larger herd sizes and increased intensification in the commercial pig industry may promote endemic infection; however, the impact that intensification has on swIAV infection dynamics and evolution is unclear. We carried out a longitudinal surveillance study for over 18 months on two enzootically infected, intensive, indoor, and multi-site pig production flows. Frequent sampling of all production stages using individual and group sampling methods was performed, followed by virological and immunological testing and whole-genome sequencing. We identified weaned pigs between 4 and 12-weeks old as the main reservoir of swIAV in the production flows, with continuous, year-round infection. Despite the continuous nature of viral circulation, infection levels were not uniform, with increasing exposure at the herd level associated with reduced viral prevalence followed by subsequent rebound infection. A single virus subtype was maintained on each farm for the entire duration of the study. Viral evolution was characterised by long periods of stasis punctuated by periods of rapid change coinciding with increasing exposure within the herd. An accumulation of mutations in the surface glycoproteins consistent with antigenic drift was observed, in addition to amino acid substitutions in the internal gene products as well as reassortment exchange of internal gene segments from newly introduced strains. These data demonstrate that long-term, continuous infection of herds with a single subtype is possible and document the evolutionary mechanisms utilised to achieve this.

Keywords: infection dynamics; intensive pig production; swine influenza A virus; viral evolution.

Figure 1.

Overview of sampling strategy, production flow organisation, and management practices. (Left) Sample types and number of samples collected per production stage per visit. (Middle) Production stages showing progression milestones from farrow to slaughter. (Right) Production flow organisation including population at each production stage, distances between sites and management practices.

Figure 2.

Comparison of viral prevalence with antibody status. Nasal wipes, udder wipes, and oral fluids were tested for the presence of swIAV by RRT-PCR and plotted as percentage positivity (left y-axis, blue bars). Each data point for gilts, weaners, and finishers represents 15 samples (10 nasal wipes and 5 oral fluids), except for the initial January 2021 time-point that represents 12 samples (10 nasal wipes and 2 oral fluids). For sucklers, each data point represents 10 udder wipes (10 litters). Oral fluid samples were also tested for the presence of influenza NP antibodies by a blocking ELISA and plotted as S/N ratios on the same graph (right y-axis, red line). S/N ratios ≤0.6 are considered antibody positive. The dashed yellow line indicates the 0.6 S/N ratio cut-off value.

Figure 3.

Subtyping and phylogenetic analysis of swIAV. (A) Stacked bar charts showing the proportional contribution of different swIAV subtypes to the overall prevalence for each production stage on both flows: pH1N1 flow 1 strains (orange), pH1N1 flow 2 strain (purple), H1N2 (green), and untyped (grey). (B) Maximum likelihood phylogenetic tree of representative pH1N1 and H1N2 sequences obtained based on complete HA coding sequences, with 1,000 bootstrap replications. Trees colour-coded according to the subtype/ strain: pH1N1 flow 1 strains (orange), pH1N1 flow 2 strain (purple), H1N2 (green), reference swine and human sequences (black), swine vaccine strains (blue).

Figure 4.

Amino acid substitutions in the swIAV surface glycoproteins over time. (A) Amino acid substitutions in HA of flow 1 weaner unit pH1N1 clade 1A.3.3.2 strains. Consensus sequences were derived from strains isolated during the first 6 months of the study. Deviations from the consensus are highlighted. Grey highlights indicate substitutions that were detected during a single time-point. Red highlights represent substitutions detected during two or more time-points. HA numbering is according to HA0, following cleavage of the 17-residue signal sequence. (B) HA homology model of A/swine/Northern Ireland/2012- 08/2021 (flow 1 weaner unit pH1N1 clade 1A.3.3.2, collection date April 2021). HA1 domain (cyan) and HA2 domain (green) from a single monomer is highlighted with locations of major substitutions indicated in red. (C) Surface rendering of A/swine/Northern Ireland/2012-08/2021 HA homology model. Antigenic sites indicated: Sa (blue), Sb (pink), Ca1 (orange), Ca2 (green), and Cb (purple). Locations of major substitutions highlighted in red. (D) Amino acid substitutions in HA of flow 2 weaner unit H1N2 clade 1B.2 strains. (E) HA homology models of A/swine/Northern Ireland/2006-10/2021 (flow 2 weaner unit H1N2 clade 1B.2, collection date March 2021). HA1 domain (cyan) and HA2 domain (green) from a single monomer is highlighted with locations of major substitutions indicated in red. (F) Surface rendering of A/swine/Northern Ireland/2006-10/2021 HA homology model. Antigenic sites indicated: Sa (blue), Sb (pink), Ca1 (orange), Ca2 (green), and Cb (purple). Locations of major substitutions highlighted in red. The location of buried residues, not visible in surface renderings, is indicated by a dashed circle. (G) NA homology model of A/swine/Northern Ireland/2006-10/2021. Mem5 epitope (cyan), monoclonal antibody escape mutations (blue). Locations of major substitutions indicated in red.

Figure 5.

Evolution of the six internal gene segments. Amino acid substitutions in the major products of the six internal gene segments of (A) flow 1 weaner unit pH1N1strains and (B) flow 2 weaner unit H1N2 strains. Consensus sequences were derived from strains isolated during the first 6 months of the study. Deviations from the consensus are highlighted. Grey highlights indicate substitutions that were detected during a single time-point. Red highlights represent substitutions detected during two or more time-points. Purple blocks indicate reassortment of segments. (C) Maximum likelihood phylogenetic tree of all pH1N1 and H1N2 sequences obtained based on complete PB2 coding sequences, with 1,000 bootstrap replications. Trees colour-coded according to the sub-type/strain: pH1N1 flow 1 strains (orange), pH1N1 flow 2 strain (purple), and H1N2 (green). Arrow indicates a reassortant H1N2, containing internal gene segments from pH1N1 flow 1 strains. (D) Schematic of the internal gene constellations of all sequenced strains.